The use of computed tomography for the detection of gastrointestinal injury in patients with penetrating torso trauma: a central South African experience

Abstract

Background: There is a consensus that selective nonoperative management is the strategy of choice for patients with penetrating torso trauma that are haemodynamically stable and do not have generalised peritonitis. Amongst two prominent trauma associations Computed Tomography is either considered part of the management strategy or given as an option depending on available resources. However, during a systematic review conducted in 2018 CT was found to fall short in the evaluation of possible gastrointestinal injury with a prevalence of 8,7% of false negative CT scans. Given that in our clinical setting there is a high incidence of penetrating torso trauma it is of value to review our local accuracy in diagnosing penetrating gastrointestinal injury using CT and to determine which direct and/or indirect signs are of greatest value for doing so. Objectives: Determining the local accuracy for identifying gastrointestinal injury on CT and reviewing the local sensitivity and specificity of some described signs for the detection of GIT injury. Method CT scans that included the torso and clinical records were reviewed for all patients with penetrating torso trauma that presented to Pelonomi Academic Hospital trauma centre from 18 December 2017 until 18 June 2019 and received a CT scan prior to surgery or discharge after a minimum of 24 hours serial clinical examination. A select number of direct and indirect CT signs for gastrointestinal injury were documented and compared to surgically confirmed GIT injury or presumed negative for GIT injury if the patient was discharged after 24 hours serial clinical examinations without clinical features suggestive of GIT injury. Results: Of the 114 cases that formed part of the study 44 (38,6%) had surgically proven GIT injury. GIT wall thickening and intra peritoneal free fluid at a site distant to the tract in the absence of solid organ injury had the greatest sensitivity and specificity of the signs evaluated each with a sensitivity of 77,3% and 73,5% and a specificity of 82,9 and 89,7% respectively. In spite of the overall high local sensitivity of 93,2% for identifying GIT injury based on CT findings, there was a prevalence of 6,8% of false negative cases. Conclusion: CT may be used as an adjunct in the evaluation of patients with penetrating torso trauma with clinical evaluation as the mainstay of management as CT cannot safely exclude GIT injury.